Cortical electrical stimulation can improve motor function after stroke, but which regions should be targeted for maximal recovery is not known. A human clinical trial of epidural motor cortex stimulation after stroke failed, but post hoc analysi revealed that the treatment was effective when targeted to spared corticospinal circuits. The critical cortical regions for restoration of motor function are being identified by studies of spontaneous recovery of function in various stroke models. These data lead to the hypothesis that the motor cortex regions that enable spontaneous recovery of motor function differ depending on the location of the stroke and that targeting electrical stimulation to these regions will produce the greatest recovery of function. To test this hypothesis, the proposed study will employ rat models of internal capsule and brain stem strokes, the most common forms of subcortical stroke in humans. Aim 1 seeks to identify the circuits responsible for spontaneous recovery in each of the models. Forelimb motor function, including a novel measure of supination, will be measured for 6 weeks after lesion to quantify spontaneous motor recovery. Adaptation of spared descending motor pathways to injury will be assayed with anatomical tract tracing and physiological motor mapping. The cortical region which enables spontaneous recovery of the impaired forelimb is hypothesized to be different for the two lesions: hindlimb motor cortex after internal capsule infarction and forelimb motor cortex after a pyramidotomy, through a brain stem bypass circuit. The necessity of these pathways for motor recovery will be tested using a viral technique that selectively and reversibly inactivates each of the circuits tha enable recovery. Aim 2 seeks to determine the efficacy of forelimb versus hindlimb motor cortex stimulation on recovery of forelimb motor function in rats with chronic subcortical stroke. For the first time, the efficacy of on- target versus off-target motor cortex stimulation will be tested. Epidural motor cortex stimulation will be applied 4 weeks after injury. The effectiveness of stimulation is hypothesized to be opposite for the two lesion types: hindlimb motor cortex stimulation will produce greater motor recovery after internal capsule infarction, and forelimb stimulation will be best after pyramidotomy. Finally, Aim 3 will test the hypothesis that motor cortex electrical stimulation will selectively strengthen the pathways that mediate spontaneous recovery, using the same anatomy, physiology, and inactivation methods as in Aim 1. The proposed research is highly innovative because of the approach-comparing different stimulation locations for different lesion types-and because it employs advanced behavioral and viral tools to identify and inactivate discrete descending motor pathways. By targeting spared connections, the therapy could be applied to many of the 7 million people living with chronic stroke. Understanding the effects of targeted epidural stimulation will enable the design of more effective protocols to restore arm and hand function in people with subcortical stroke and persistent motor impairment.